Yesterday in part two of our series on Derek Muller (Veritasium)'s Facebook Like Fraud video, we covered Derek's own experiences with advertising through Facebook on the Veritasium Facebook page, explained why his ads didn't perform well, and how you could avoid making those same mistakes. Today, we're going to examine Virtual Cat, his follow up experiment that expands on the Virtual Bagel experiment by BBC journalist Rory Cellan-Jones.
This post ended up a bit long, so I've broken it up into two parts. In this first half, we review the Virtual Cat Facebook page ad experiment, and in the second half we take a deeper look at the Virtual Cat Facebook page, and whether Derek's conclusions based on this experiment really make sense. You can read the second half of this blog post here.
Virtual Bagel, V.2?
At first, Virtual Cat looks like a rehashing of Virtual Bagel, just 2 years after the original. Derek creates a fake page that obviously serves no real purpose, runs a small ad campaign, and examines the resulting engagement of the likes he courted. However, Virtual Cat differs in three distinct ways:
1. The ads are targeted only to cat lovers in the US, Canada, Australia, and the UK.
2. Rather than pose as a silly and sarcastic business page that some people might "like" simply because the premise is amusing, Virtual Cat's page description actually insults the people who like the page, stating "Only an idiot would like this page."
3. While Derek certainly drew some conclusions out of this short lived experiment, he clearly never examined how the audience he built would engage with content. Just a day after it was created, Veritasium fans are all up in Derek's experiment, tainting the results.
So, Derek had expected that by excluding the countries that generally provide low-quality likes that he would see an increase in engagement and only high-quality, valuable likes. He spent $10 on Facebook ads and gained 39 likes within 20 minutes at a cost of $0.20 per click. However, he was disappointed with his results, as he found the liking habits of his new fans as odd. They liked many hundreds of pages, and sometimes liked multiple brands that were direct competitors. That's pretty much the extent of the evidence Derek discusses in his video regarding Virtual Cat before moving on to sharing hypotheses as to why this might be.
Whoa, Did We Just Skip the Whole Point of the Experiment?
The original problem Derek saw in advertising with the Veritasium Facebook page was that engagement decreased as the number of likes increased, but he doesn't mention how engagement fared on the Virtual Cat page as it grew, so I had to look into it myself. However, upon doing that, I found that from the page's inception it was swamped by Veritasium fans. Then I remember that Derek mentioned in his video that he set the page up "yesterday," and I realize that this experiment didn't go on for very long at all. He never really tried to gauge engagement from these fans, judging them instead based on the other likes of the few he could actually examine. Which, I agree those are some interesting behaviors, but what do they really mean? We don't know if those fans will engage or not, because apparently that (bafflingly) wasn't part of the experiment.
Liking habits that don't fit how you or people you know like pages doesn't really prove anything. What matters, just like with the Virtual Bagel experiment, and just like with the Veritasium Facebook page, is whether those new fans engage with the content in a meaningful way. I can't understand why Derek's first two examples focus on the importance of engagement, but then the Virtual Cat experiment skips over it entirely, nor why he did the experiment the day before publishing his video, offering no opportunity to accurately measure how this group would engage with content before the page became inundated by people from other sources who had watched the video after it went live.
This post ended up being quite long, so I've broken it up into two posts. Click through to read the second half of this post, where I examine the method and results of Derek's Virtual Cat experiment in more detail!